Friday, January 25, 2013

Customer Analytics: Custom-fit, timely ans just a bit creepy!

Customer Analytics:
Custom-fit, timely and just a bit creepy!

     You're browsing the web, searching for hotels in San Francisco. The next time you come to your search page, there are deal offers for flights from your town to SFO. On your email account there are great coupons for a restaurant on Fisherman's Wharf. Wait! Discount tickets for a tour of Alcatraz! Hold on a moment ... who knows I'm going to a conference in the Bay Area? The Web knows. Sites you have visited are leveraging customer analytics to decide what to show you.

     As with any technology, service - or any aspect of life for that matter - there are always risks and benefits to this form of marketing.  The long and short of it is that customer analytics differ from other forms of analytics in that they deal with specific opinions, feedback and behavior of individuals, not visitors as a group. Sites and services are interested in what their individual customers are doing.  They will then use this information to improve their sites or to individualize a singe customer's experience, ads, etc.

     At the risk of paraphrasing a method actor, what's the motivation here?The general motivation and desire of any website is to get the visitor to act in the intended way based upon that site's mission.  If it is a travel site, they want  you to book travel, a strictly informative site wants you to read their content, a search site wants you to click on their paid advertisers and find what you are looking for quickly, an eCommerce site wants you to buy. Beyond the simple idea of getting you to buy, they want you to stay on their site, find what you are looking for and feel happy with the experience so that you will come back
     The first step for sites to do is to identify an individual. This is primarily done using cookies. A site assigns an identity which is stored in a cookie. This allows a site to track an individual browser, and presumably, a unique customer.  This information will be lost if the user deletes cookies.  The information will also be less accurate if multiple people share the computer/browser (it's a bit less than ideal for a kid to get ads for viagra when he wanted to look at action figures). Ideally the site will be able to entice a visitor to sign on to the site, make a credit card purchase, sign up for a loyalty account, or use some other function that more truly uniquely identifies him or her to the site.  Furthermore, once someone signs in, their demographic information can be stored and associated with previous matched visits (by the cookie identifier) and data regarding this person's activity can be collected and acted upon.
     What are the things that can be captured?

  • Purchases
  • Product views
  • Shopping cart additions
  • Video views
  • Review reads
  • Registering for a webinar
  • Accessing downloads
  • Requesting help
  • Forwarding a link
  • Posting a comment
  • GPS location - some mobile devices
     Individual information from customers can also be obtained through online surveys and interactive chat.
     It must be remembered that with all web analysis information, data can go stale. A person's travel interests will not necessarily be the same in six months. Also, once a big screen tv or car is purchased, a person will start to get annoyed by more ads regarding these products.

Some uses of these data are successful:
  • Market deals on similar cheaper items when items in a cart are abandoned
  • Market region-specific deals
  • Market products, deals, etc. after a search and before a purchase
Conversely, some will be failures:
  • Continue to market products after a purchase is made
  • Incorrectly identify a home city and market travel from the wrong city
On the other hand, some might find the degree of information collected to be an invasion of privacy.

A bit of an extreme illustration
from SuperNews

     Business will always do as Business does. If there is a business technology that serves to increase success and profitability, this route will be pursued (at least where legal). Some degree of security can be maintained by purging cookies every time the browser is exited. Features like In Private Browsing can help with this. On the other hand, if you sign on to a website to use all its features, it is going to be a little tougher to maintain privacy.
     Clearly there are advantages to the use of these analytics, people do buy, act and respond to the directed marketing. Furthermore, many must appreciate it or it would not be successful and would soon disappear.



  1. Love the video! It definitely feels like that's the direction things are heading! I'm ok with companies trying to measure how they can better serve my needs, as long as it's only to a certain extent and not take too far. Nice post!

  2. Great post, Gordon. Reminds me of a story I read last year...A young girl bought prenatal vitamins and some other stuff at Target, and a while later Target mailed her a little catalog full of pregnancy stuff. Her dad freaked out at Target and was like, "This is inappropriate for her, what's wrong with you?" Then had to apologize when he found out that Target's customer analytics learned of his daughter's pregnancy before he did.

  3. I love the first paragraph. It totally relates to my personal experience with highly relevant webpage ads.